Best AI for business emails and everyday work: all 77 ranked models
Ordered by business-task votes. Score shown: how often people preferred its answers to business questions. Higher is better, and neighbouring models are often too close to separate: check the range before reading much into a one-place gap.
This is one test, not a recommendation. Our pick for get work done also looks at following-instructions votes and everyday chat votes, and flags close calls. See the options worth considering.
| # | Model | Position | Score | Rating | Source |
|---|---|---|---|---|---|
| 1 | Gemini 4 Argon Best-supported choiceGoogle DeepMind | Ranked 1st of 77 | 1,51595% range 1,502 to 1,529 | Top tier | Source for Gemini 4 Argon get work done score (lmarena) |
| 2 | Claude Opus 5 Good-enough alternativeAnthropic | Ranked 2nd of 77 | 1,50095% range 1,492 to 1,508 | Top tier | Source for Claude Opus 5 get work done score (lmarena) |
| 3 | Claude Opus 5.5 Good-enough alternativeAnthropic | Ranked 3rd of 77 | 1,49795% range 1,483 to 1,511 | Top tier | Source for Claude Opus 5.5 get work done score (lmarena) |
| 4 | Claude Opus 4.6Anthropic | Ranked 4th of 77 | 1,49795% range 1,491 to 1,503 | Top tier | Source for Claude Opus 4.6 get work done score (lmarena) |
| 5 | Claude Fable 5.1Anthropic | Ranked 5th of 77 | 1,49495% range 1,483 to 1,505 | Top tier | Source for Claude Fable 5.1 get work done score (lmarena) |
| 6 | Claude Opus 4.7Anthropic | Ranked 6th of 77 | 1,49095% range 1,484 to 1,497 | Top tier | Source for Claude Opus 4.7 get work done score (lmarena) |
| 7 | Muse WonderMeta | Ranked 7th of 77 | 1,48795% range 1,471 to 1,504 | Top tier | Source for Muse Wonder get work done score (lmarena) |
| 8 | Muse Spark 1.3Meta | Ranked 8th of 77 | 1,48695% range 1,476 to 1,496 | Top tier | Source for Muse Spark 1.3 get work done score (lmarena) |
| 9 | Muse Spark 1.2Meta | Ranked 9th of 77 | 1,48495% range 1,470 to 1,497 | Top tier | Source for Muse Spark 1.2 get work done score (lmarena) |
| 10 | Qwen3.8-MaxAlibaba | Ranked 10th of 77 | 1,48095% range 1,472 to 1,489 | Top tier | Source for Qwen3.8-Max get work done score (lmarena) |
| 11 | Gemini 3.7 FlashGoogle DeepMind | Ranked 11th of 77 | 1,48095% range 1,472 to 1,489 | Top tier | Source for Gemini 3.7 Flash get work done score (lmarena) |
| 12 | Gemini 3.8 FlashGoogle DeepMind | Ranked 12th of 77 | 1,48095% range 1,473 to 1,487 | Top tier | Source for Gemini 3.8 Flash get work done score (lmarena) |
| 13 | Claude Fable 5Anthropic | Ranked 13th of 77 | 1,48095% range 1,473 to 1,487 | Top tier | Source for Claude Fable 5 get work done score (lmarena) |
| 14 | GPT-5.5OpenAI | Ranked 14th of 77 | 1,47895% range 1,472 to 1,484 | Top tier | Source for GPT-5.5 get work done score (lmarena) |
| 15 | GPT-5.4OpenAI | Ranked 15th of 77 | 1,47595% range 1,468 to 1,481 | Top tier | Source for GPT-5.4 get work done score (lmarena) |
Show the other 62 ranked models
| # | Model | Position | Score | Rating | Source |
|---|---|---|---|---|---|
| 16 | ERNIE 5.1Baidu | Ranked 16th of 77 | 1,47395% range 1,465 to 1,481 | Top tier | Source for ERNIE 5.1 get work done score (lmarena) |
| 17 | Muse Spark 1.1Meta | Ranked 17th of 77 | 1,47095% range 1,463 to 1,478 | Top tier | Source for Muse Spark 1.1 get work done score (lmarena) |
| 18 | Kimi K3 Best open-weights option Open weightsMoonshot | Ranked 18th of 77 | 1,46995% range 1,461 to 1,477 | Top tier | Source for Kimi K3 get work done score (lmarena) |
| 19 | MiMo-V2.6-Pro Open weightsXiaomi | Ranked 19th of 77 | 1,46895% range 1,446 to 1,490 | Top tier | Source for MiMo-V2.6-Pro get work done score (lmarena) |
| 20 | Gemini 3.5 FlashGoogle DeepMind | Ranked 20th of 77 | 1,46795% range 1,460 to 1,474 | Top tier | Source for Gemini 3.5 Flash get work done score (lmarena) |
| 21 | MiMo-V2.5-Pro Open weightsXiaomi | Ranked 21st of 77 | 1,46795% range 1,461 to 1,472 | Strong | Source for MiMo-V2.5-Pro get work done score (lmarena) |
| 22 | Claude Sonnet 5.5Anthropic | Ranked 22nd of 77 | 1,46695% range 1,451 to 1,481 | Strong | Source for Claude Sonnet 5.5 get work done score (lmarena) |
| 23 | Gemini 3.6 FlashGoogle DeepMind | Ranked 23rd of 77 | 1,46495% range 1,456 to 1,472 | Strong | Source for Gemini 3.6 Flash get work done score (lmarena) |
| 24 | Claude Sonnet 4.6Anthropic | Ranked 24th of 77 | 1,46495% range 1,458 to 1,470 | Strong | Source for Claude Sonnet 4.6 get work done score (lmarena) |
| 25 | Qwen3.5-Max-PreviewAlibaba | Ranked 25th of 77 | 1,46395% range 1,454 to 1,472 | Strong | Source for Qwen3.5-Max-Preview get work done score (lmarena) |
| 26 | Muse SparkMeta | Ranked 26th of 77 | 1,46295% range 1,451 to 1,473 | Strong | Source for Muse Spark get work done score (lmarena) |
| 27 | DeepSeek-V4.1-Flash-20260910 Open weightsDeepSeek | Ranked 27th of 77 | 1,46295% range 1,450 to 1,474 | Strong | Source for DeepSeek-V4.1-Flash-20260910 get work done score (lmarena) |
| 28 | GLM-5.3-Flash Open weightsZ.ai (Zhipu AI) | Ranked 28th of 77 | 1,46195% range 1,453 to 1,469 | Strong | Source for GLM-5.3-Flash get work done score (lmarena) |
| 29 | GLM-5.3 Open weightsZ.ai (Zhipu AI) | Ranked 29th of 77 | 1,46195% range 1,452 to 1,470 | Strong | Source for GLM-5.3 get work done score (lmarena) |
| 30 | Gemini 3.1 Pro PreviewGoogle DeepMind | Ranked 30th of 77 | 1,45995% range 1,454 to 1,464 | Strong | Source for Gemini 3.1 Pro Preview get work done score (lmarena) |
| 31 | Claude Opus 4.8Anthropic | Ranked 31st of 77 | 1,45895% range 1,452 to 1,465 | Strong | Source for Claude Opus 4.8 get work done score (lmarena) |
| 32 | Gemini 3 ProGoogle DeepMind | Ranked 32nd of 77 | 1,45895% range 1,451 to 1,465 | Strong | Source for Gemini 3 Pro get work done score (lmarena) |
| 33 | GPT-5.6 SolOpenAI | Ranked 33rd of 77 | 1,45695% range 1,449 to 1,464 | Strong | Source for GPT-5.6 Sol get work done score (lmarena) |
| 34 | GLM-5.1 Open weightsZ.ai (Zhipu AI) | Ranked 34th of 77 | 1,45395% range 1,447 to 1,460 | Strong | Source for GLM-5.1 get work done score (lmarena) |
| 35 | GLM-5.2 Open weightsZ.ai (Zhipu AI) | Ranked 35th of 77 | 1,45195% range 1,444 to 1,457 | Strong | Source for GLM-5.2 get work done score (lmarena) |
| 36 | MiMo-V2.6-Flash Open weightsXiaomi | Ranked 36th of 77 | 1,45095% range 1,439 to 1,462 | Strong | Source for MiMo-V2.6-Flash get work done score (lmarena) |
| 37 | amazon-nova-experimental-chat (26-02-10)Amazon | Ranked 37th of 77 | 1,45095% range 1,428 to 1,472 | Strong | Source for amazon-nova-experimental-chat (26-02-10) get work done score (lmarena) |
| 38 | GPT-6.1 SolOpenAI | Ranked 38th of 77 | 1,44995% range 1,433 to 1,464 | Strong | Source for GPT-6.1 Sol get work done score (lmarena) |
| 39 | Claude Opus 4.5 (20251101)Anthropic | Ranked 39th of 77 | 1,44895% range 1,442 to 1,454 | Strong | Source for Claude Opus 4.5 (20251101) get work done score (lmarena) |
| 40 | Gemini 3 FlashGoogle DeepMind | Ranked 40th of 77 | 1,44795% range 1,439 to 1,455 | Middle | Source for Gemini 3 Flash get work done score (lmarena) |
| 41 | GPT-5.6 TerraOpenAI | Ranked 41st of 77 | 1,44795% range 1,440 to 1,454 | Middle | Source for GPT-5.6 Terra get work done score (lmarena) |
| 42 | Qwen3.7-Max-PreviewAlibaba | Ranked 42nd of 77 | 1,44695% range 1,426 to 1,466 | Middle | Source for Qwen3.7-Max-Preview get work done score (lmarena) |
| 43 | Kimi K2.6 Open weightsMoonshot | Ranked 43rd of 77 | 1,44595% range 1,438 to 1,452 | Middle | Source for Kimi K2.6 get work done score (lmarena) |
| 44 | DeepSeek-V4-Pro Open weightsDeepSeek | Ranked 44th of 77 | 1,44395% range 1,436 to 1,450 | Middle | Source for DeepSeek-V4-Pro get work done score (lmarena) |
| 45 | DeepSeek-V4-Pro (High, 20260813) Open weightsDeepSeek | Ranked 45th of 77 | 1,44295% range 1,428 to 1,456 | Middle | Source for DeepSeek-V4-Pro (High, 20260813) get work done score (lmarena) |
| 46 | Grok 4.5xAI | Ranked 46th of 77 | 1,44195% range 1,434 to 1,449 | Middle | Source for Grok 4.5 get work done score (lmarena) |
| 47 | Gemini 2.5 ProGoogle DeepMind | Ranked 47th of 77 | 1,43995% range 1,435 to 1,444 | Middle | Source for Gemini 2.5 Pro get work done score (lmarena) |
| 48 | GPT-5.6 LunaOpenAI | Ranked 48th of 77 | 1,43995% range 1,432 to 1,446 | Middle | Source for GPT-5.6 Luna get work done score (lmarena) |
| 49 | Dola Seed 2.0 ProByteDance | Ranked 49th of 77 | 1,43995% range 1,433 to 1,445 | Middle | Source for Dola Seed 2.0 Pro get work done score (lmarena) |
| 50 | Mistral Large 4Mistral AI | Ranked 50th of 77 | 1,43995% range 1,425 to 1,452 | Middle | Source for Mistral Large 4 get work done score (lmarena) |
| 51 | GPT-6 AstraOpenAI | Ranked 51st of 77 | 1,43895% range 1,426 to 1,450 | Middle | Source for GPT-6 Astra get work done score (lmarena) |
| 52 | Claude Sonnet 5Anthropic | Ranked 52nd of 77 | 1,43795% range 1,430 to 1,444 | Middle | Source for Claude Sonnet 5 get work done score (lmarena) |
| 53 | NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4) Open weightsNVIDIA | Ranked 53rd of 77 | 1,43595% range 1,422 to 1,448 | Middle | Source for NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4) get work done score (lmarena) |
| 54 | Qwen3.7-PlusAlibaba | Ranked 54th of 77 | 1,43595% range 1,427 to 1,442 | Middle | Source for Qwen3.7-Plus get work done score (lmarena) |
| 55 | Qwen3.5-397B-A17B Open weightsAlibaba | Ranked 55th of 77 | 1,43595% range 1,429 to 1,440 | Middle | Source for Qwen3.5-397B-A17B get work done score (lmarena) |
| 56 | Grok 4.20 Beta (Reasoning, 0309)xAI | Ranked 56th of 77 | 1,43295% range 1,426 to 1,438 | Middle | Source for Grok 4.20 Beta (Reasoning, 0309) get work done score (lmarena) |
| 57 | GPT-5.1OpenAI | Ranked 57th of 77 | 1,43195% range 1,424 to 1,438 | Middle | Source for GPT-5.1 get work done score (lmarena) |
| 58 | Gemma 4 31B Open weightsGoogle DeepMind | Ranked 58th of 77 | 1,43095% range 1,414 to 1,447 | Middle | Source for Gemma 4 31B get work done score (lmarena) |
| 59 | GLM-5 Open weightsZ.ai (Zhipu AI) | Ranked 59th of 77 | 1,43095% range 1,422 to 1,438 | Behind the leaders | Source for GLM-5 get work done score (lmarena) |
| 60 | Qwen3.6-PlusAlibaba | Ranked 60th of 77 | 1,43095% range 1,423 to 1,437 | Behind the leaders | Source for Qwen3.6-Plus get work done score (lmarena) |
| 61 | GLM-4.7 Open weightsZ.ai (Zhipu AI) | Ranked 61st of 77 | 1,42995% range 1,417 to 1,441 | Behind the leaders | Source for GLM-4.7 get work done score (lmarena) |
| 62 | ERNIE 5.0 (0110)Baidu | Ranked 62nd of 77 | 1,42995% range 1,421 to 1,436 | Behind the leaders | Source for ERNIE 5.0 (0110) get work done score (lmarena) |
| 63 | DeepSeek-V4-Pro Preview Open weightsDeepSeek | Ranked 63rd of 77 | 1,42895% range 1,421 to 1,434 | Behind the leaders | Source for DeepSeek-V4-Pro Preview get work done score (lmarena) |
| 64 | Qwen3.6-Max-PreviewAlibaba | Ranked 64th of 77 | 1,42795% range 1,409 to 1,446 | Behind the leaders | Source for Qwen3.6-Max-Preview get work done score (lmarena) |
| 65 | Kimi K2.5 Open weightsMoonshot | Ranked 65th of 77 | 1,42795% range 1,421 to 1,432 | Behind the leaders | Source for Kimi K2.5 get work done score (lmarena) |
| 66 | Grok 4.20 Beta 1xAI | Ranked 66th of 77 | 1,42595% range 1,416 to 1,433 | Behind the leaders | Source for Grok 4.20 Beta 1 get work done score (lmarena) |
| 67 | Gemini 3.5 Flash-LiteGoogle DeepMind | Ranked 67th of 77 | 1,42395% range 1,415 to 1,431 | Behind the leaders | Source for Gemini 3.5 Flash-Lite get work done score (lmarena) |
| 68 | Grok 4.20 Multi-Agent Beta (0309)xAI | Ranked 68th of 77 | 1,42395% range 1,416 to 1,429 | Behind the leaders | Source for Grok 4.20 Multi-Agent Beta (0309) get work done score (lmarena) |
| 69 | Grok 4.6xAI | Ranked 69th of 77 | 1,41695% range 1,408 to 1,424 | Behind the leaders | Source for Grok 4.6 get work done score (lmarena) |
| 70 | Inkling-Small Open weightsThinking Machines | Ranked 70th of 77 | 1,41495% range 1,406 to 1,422 | Behind the leaders | Source for Inkling-Small get work done score (lmarena) |
| 71 | GPT-5.5 InstantOpenAI | Ranked 71st of 77 | 1,41095% range 1,401 to 1,418 | Behind the leaders | Source for GPT-5.5 Instant get work done score (lmarena) |
| 72 | Claude Haiku 5.5Anthropic | Ranked 72nd of 77 | 1,40595% range 1,389 to 1,422 | Behind the leaders | Source for Claude Haiku 5.5 get work done score (lmarena) |
| 73 | Qwen3.5-FlashAlibaba | Ranked 73rd of 77 | 1,39495% range 1,388 to 1,401 | Behind the leaders | Source for Qwen3.5-Flash get work done score (lmarena) |
| 74 | Grok 4.7xAI | Ranked 74th of 77 | 1,39095% range 1,378 to 1,402 | Behind the leaders | Source for Grok 4.7 get work done score (lmarena) |
| 75 | GPT-6 LunaOpenAI | Ranked 75th of 77 | 1,38495% range 1,378 to 1,391 | Behind the leaders | Source for GPT-6 Luna get work done score (lmarena) |
| 76 | GPT-6 SolOpenAI | Ranked 76th of 77 | 1,38195% range 1,369 to 1,393 | Behind the leaders | Source for GPT-6 Sol get work done score (lmarena) |
| 77 | Grok 4.3xAI | Ranked 77th of 77 | 1,38195% range 1,375 to 1,387 | Behind the leaders | Source for Grok 4.3 get work done score (lmarena) |
Test: Arena text: business and management. Source: lmarena. Data last changed 29 Oct 2026. How we test.
Want a straight answer for get work done?
See the options the evidence supports, with limits and verified prices.